The Joint Product Trap

An investigation that starts with bad packaging and ends with the structure of global trade.

You have opened a package of bacon. The experience is familiar and universally disliked. The vacuum-sealed plastic window resists your fingers. You reach for scissors, or a knife, or you just pull until something tears unevenly. Once open, the package cannot be resealed. The remaining strips sit in the refrigerator, exposed, accelerating their own spoilage. You think: surely someone could design a better package than this.

Someone could. Zip-lock bacon packs exist. Rigid tubs with snap lids exist. They add 30 to 50 cents per unit. At a price point where consumers demonstrably switch brands for a quarter, that is not a rounding error — it is a competitive disadvantage that can move market share. Consumer surveys consistently rank Bacon packaging among the worst in the grocery store. Yet it persists, unchanged, decade after decade.

The package is not designed for you. It is designed for the packer, who needs vacuum sealing to extend shelf life from days to weeks. For the shipper, who needs the flat format to stack efficiently in cold chain containers. For the retailer, who needs the transparent window so shoppers can evaluate lean-to-fat ratio and the form factor to fit standard meat case planograms. You are the last node in this chain and the least powerful, so you get whatever is left over after everyone upstream has been served.

This pattern has a name: supply-chain-optimized design. The end user's experience is the residual outcome of upstream optimization. You see it in bacon packaging, pharmaceutical blister packs, airline seat pitch, and enterprise software. The design is not broken. It is optimized — just not for you.

Why is the bacon itself expensive?

But the packaging frustration leads somewhere more interesting. If the supply chain optimizes everything upstream and passes the residual downstream, what does that tell you about the price of the bacon inside the package?

The average price of bacon in the US has more than doubled since 2010. It is not driven by packaging costs, or labor, or transportation. It is driven by the fact that a pork belly — the anatomical cut from which bacon is made — is a scarce component drawn from an animal that produces it in a fixed biological proportion, and that component is simultaneously desired by buyers all over the world in forms that have nothing to do with American breakfast.

The belly constitutes 8 to 10 percent of a hog's carcass by weight, but represents 15 to 19 percent of total carcass value because demand disproportionately concentrates on it. A single hog yields exactly one set of bellies. There is no belly mine. You cannot dig deeper. If you want more bellies, you must slaughter more hogs, which produces more of every other cut simultaneously.

The animal as a fixed bundle

This is the core of the problem, and it is worth stating precisely. A market hog yields a carcass of roughly 210 pounds, broken into primal cuts — bellies, loins, hams, shoulders, ribs, trim, offal — in proportions determined by genetics and feeding. No packer can change these proportions at processing time. The hog is a fixed bundle of joint products.

This means the marginal cost of one additional pound of pork belly is not the cost of producing a pound of belly. It is the cost of producing an entire additional hog, minus the revenue from every other cut. When belly demand rises, the rational response — slaughter more hogs — floods other markets, depresses loin and ham prices, reduces the offset revenue, and raises the effective marginal cost of bellies. The system has a built-in negative feedback loop. Economists call this the joint product problem: when a single input yields multiple outputs in fixed proportions, you cannot change the supply of one without changing the supply of all the others. The trap is that every lever you pull moves five other variables in directions you do not want.

A decision to breed more sows today produces market hogs in 10 to 11 months. The pork supply chain responds to price signals with nearly a year of lag. This is the well-documented hog cycle: high prices trigger expansion, expansion triggers oversupply, oversupply triggers contraction, contraction triggers scarcity, scarcity triggers high prices, and around it goes.

Who else wants the belly?

In the United States, most pork bellies become bacon. Globally, the belly serves a much wider set of uses, all drawing from the same anatomical cut:

Samgyeopsal — uncured, unseasoned belly slices grilled at the table — is eaten weekly by an estimated 85 percent of native Koreans. South Korea is one of the largest export destinations for US pork bellies. Mexico buys over 250 million pounds of US pork per month. Japan is another major buyer. When these export markets surge, domestic belly supply tightens, and the price of your breakfast bacon rises.

The restaurant sector adds another layer of competition. Approximately 80 percent of the top 500 US restaurant chains offer at least one bacon item. Meanwhile, fresh pork belly as a premium entrée has grown nearly 60 percent in menu penetration since 2014. These two use cases compete directly for the same raw material, with restaurants often able to outbid commodity bacon processors on a per-pound basis.

Where does the money go?

The supply chain margins are surprisingly uneven.

Producers averaged near-decade-high margins in 2025, driven by lower feed costs and stable demand. Yet the sow herd contracted 2 percent year-over-year. Even with strong profitability, capital requirements, limited processing capacity, and disease risk prevent expansion. The industry is profitable but static.

Packers operate near breakeven at roughly $5 per head, running plants at 95 percent capacity to survive on volume. The real margin capture happens at the processing stage, where turning raw bellies into retail bacon commands a gross margin of $5.43 per pound — up from $4.45 in 2014. This widening spread reflects growing demand, brand premiums, and the capital intensity of modern curing and high-speed slicing operations.

Retailers frequently use bacon as a loss leader, selling it below full margin to drive store traffic. Your bacon is subsidized by the margin on the rest of your cart.

Why nobody can make it cheaper

If you have followed the logic to this point, you can see why the obvious fixes do not work.

Raise more hogs. More hogs means more of every cut, depressing prices elsewhere, reducing offset revenue, and potentially making expansion uneconomic. Even at historically high margins, producers are not expanding. The risks and capital requirements are too large.

Breed pigs with bigger bellies. Decades of breeding leaner hogs have already shrunk the belly from 15 to 18 percent to 8 to 10 percent of the carcass. Reversing this would produce fattier loins and hams that consumers do not want. You cannot optimize one region of the animal in isolation.

Reduce exports. Would increase domestic supply in the short term but undermine producer margins, trigger industry contraction, and invite retaliatory trade measures against US agricultural exports broadly.

Substitute products. Beef bacon is growing at about 5.3 percent annually, and plant-based alternatives exist, but non-pork bacon holds under 8 percent market share. Most consumers treat these as separate categories rather than true replacements.

Expand processing capacity. Addresses a real bottleneck but shifts rather than resolves the problem. More bacon processing capacity increases demand for bellies, driving up the belly primal price, which flows back to the consumer.

Every proposed intervention either runs into the joint product constraint, requires a 10-month response lag, creates tradeoffs in other cuts, or addresses a non-binding constraint. The system is in a locally stable equilibrium. Bacon is expensive because the belly is a scarce, fixed-proportion component of a multi-product animal, with globally competitive demand from uses that did not exist at scale 20 years ago. This is not a market failure. It is the market working correctly under genuine biological and structural constraints.

What generalizes

If you pull on the thread of bad bacon packaging, you eventually reach the biology of a 280-pound animal, the structure of global agricultural trade, and the mathematics of constrained optimization. The same pattern shows up far from the meat case.

Joint products and constrained optimization. You cannot reason about the price of one output from a joint production process without reasoning about all the others simultaneously. This is isomorphic to crude oil refining, semiconductor wafer binning, and any system where a single input produces multiple outputs in fixed ratios.

Supply-chain-optimized design. Whenever a consumer product seems obviously poorly designed, check whether it is actually optimized for a different node in the value chain. Often it is.

The binding constraint. The binding constraint on bacon supply is not capital, not labor, not processing capacity, and not demand. It is the biological fact that a belly is 8 to 10 percent of a hog. Every other constraint can be relaxed with money and time. This one cannot, short of genetic engineering on a generational timescale.

Revealed preference. Consumers say they want better packaging but will not pay 30 cents more for it. The bacon market is a laboratory for the divergence between stated and revealed preferences in a frequently purchased, price-sensitive commodity.

The next time you open a terrible package of bacon, you are interacting with the terminal node of a supply chain that begins with a sow in Iowa, passes through a joint production system serving Korean barbecue restaurants and Italian delis simultaneously, is priced by a global commodity market sensitive to currency fluctuations and trade policy, and is packaged in a format optimized for a cold chain logistics network you will never see.

The package is bad because making it better would cost more than you would pay. The bacon is expensive because 74.6 million hogs can only produce so many bellies, and the entire world wants them.

Some problems are not solved. They are understood.